Newts in time and space: the evolutionary history of Triturus newts at different temporal and spatial scales

different temporal and spatial scales

Espregueria Themudo, G.

Citation

Espregueria Themudo, G. (2010, March 10). Newts in time and space: the evolutionary history of Triturus newts at different temporal and spatial scales. Retrieved from

https://hdl.handle.net/1887/15062

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/15062

Note: To cite this publication please use the final published version (if applicable).

N

:

T

Espregueira Themudo, G. and J.W. Arntzen

1 National Museum of Natural History – Naturalis, PO Box 9517, 2300 RA Leiden, the Netherlands.

2 CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, Campus Agrário de Vairão, 4485-661 Vairão, Portugal.

Contents

Abstract ... 158

Introduction ... 159

Material and Methods ... 161

Sampling ... 161

Genetic Data ... 161

Ecological Data ... 162

Model selection ... 163

Results ... 164

Discussion ... 167

Acknowledgments. ... 169

References ... 170

Appendix S1 ... 174

Appendix S2 ... 175

Published in Diversity and Distributions 13 (5) 580-586 (2007).

Abstract

The newt species Triturus marmoratus and T. pygmaeus are both present in central Portugal where they have parapatric distributions. We used four genetic markers to determine which species was present in 31 populations. In the centre of the study area we found a T. marmoratus enclave. Despite small inter-population distances, hybridization is locally rare. We built several models to try to explain this distribution using environmental data. The best model, chosen by Akaike’s Information Criterion, relates the presence of T.

marmoratus with the temperature in July, the relief of the landscape and a higher use of the land for orchards. The current distribution can best be explained by T. pygmaeus expanding north and replacing T. marmoratus, the latter only persisting where ecological conditions are relatively favourable.

Keywords: Akaike’s Information Criterion, Allozymes, Amphibia, Enclaves, Mosaic Hybrid Zones, Parapatry

Introduction

Secondary contact between closely related species often generates spectacular events and provides insight into the evolutionary process. Hybrid zones formed this way are considered ‘natural laboratories’ or ‘windows on evolutionary processes’ (e.g. HEWITT, 1988 in Anartia butterflies; HARRISON, 1990). Evolution is not a static process and to understand it, we must look into dynamic systems like these.

The dynamics of species ranges are not independent of ecological conditions, competition with sister species, or dispersal capabilities. Present day distributions are echoes from past events. One particular event that provides clues to the relative movement of species is the formation of enclaves. Enclaves are here defined as populations of one species completely surrounded by populations of closely related species and genetically isolated from other populations of the same species (ARNTZEN, 1978; in geographical terms, they are simultaneously exclaves and enclaves). This is reminiscent of the ‘internal parapatry’ concept of Key (1981) that, however, does not deal with disjunct distributions.

Mosaic hybrid zones are bimodal hybrid zones with few hybrids and predominantly parental genotypes present. The contact between the species is more strongly shaped by ecological constraints than by genetic interactions. In a recent review, Jiggins & Mallet (2000) go one step further and suggest that ecology contributes more to speciation than genetic incompatibility. Well known mosaic hybrid zones are for example those in Gryllus crickets (RAND and HARRISON, 1989), Chorthippus grasshoppers (BRIDLE et al., 2001), and Mytilus mussels (BIERNE et al., 2003). Examples of mosaic hybrid zones in salamanders are Triturus cristatus and T. marmoratus in western France (ARNTZEN and WALLIS, 1991) and Plethodon cinereus and P. shenandoah in the Appalachian Mountains of North America (JAEGER, 1970; JAEGER, 1971; see also SITES et al., 2004). Although they present patches of populations of one species distributed among patches of the other, these are not all necessarily enclaves because dispersal among patches may be frequent.

To the best of our knowledge the only enclaves recorded in the literature are those for Bombina toads in central Europe (ARNTZEN, 1978) and Triturus newts in western France and the northern Balkans (ARNTZEN and WALLIS, 1991;

ARNTZEN and WALLIS, 1999). Perhaps enclaves are more likely to arise in

organisms with structured populations and low dispersal capability than in organisms that disperse well. Amphibian populations in particular are well- delimited because of their dependence on water for reproduction and they have low individual mobility. Once formed, enclaves will take some time to dissolve, or be stable or disappear by reconnection to the main distribution.

The two species of marbled newts living in the Iberian Peninsula, Triturus marmoratus (Latreille, 1800) and T. pygmaeus (Wolterstorff, 1908) have a parapatric distribution. Some reports, however, indicated the presence of T.

marmoratus where only T. pygmaeus was expected, near Caldas da Rainha and a spatial-environmental model for the two species suggests that the local conditions may indeed be favourable to T. marmoratus (see the southernmost record in Fig.

1c in ARNTZEN, 2006)). This would indicate an area of sympatry or a mosaic distribution. A mosaic distribution would point to differential ecological

requirements with patches where the environmental conditions are more suitable for one species than for the other.

An intuitive explanation from fieldwork across the Iberian Peninsula is that T. pygmaeus thrives in ephemeral ponds with a fluctuating reproductive output and that T. marmoratus thrives in smaller, more permanent water bodies (e.g., springs) with regular but low annual recruitment reaching metamorphosis (J.

W. Arntzen, unpublished). To confirm the record at Caldas da Rainha and to learn more about the events that lead to this occurrence and its extent, we conducted a detailed study. Adult marbled newts have clearly distinguishable morphologies while embryos and larvae are difficult to identify. Conversely, embryos and larvae are easy to find at the aftermath of the reproductive season while adults may be elusive. To facilitate a fast, detailed and reliable surveying we employed genetic markers for species identification (CHAPTER 7).

In the present paper, we look into the spatial structure of the distribution of the two species of marbled newts and determine if and what ecological constraints are shaping it. We also analyze the presence of hybrids in our sample.

Material and Methods Sampling

In April 2003, we searched around Caldas da Rainha, in central Portugal, for water bodies containing marbled newts. The region is characterized by flat dunes on the coast and an undulating agricultural land, with an abundance of orchards in the centre. We located and sampled 31 populations of Triturus marmoratus and T.

pygmaeus, over 1600 km². Breeding female newts obtain one or several

spermatophores from which eggs are internally fertilized and then deposited one by one, mostly on aquatic vegetation. The process takes place over a ca. two month period.

To increase random collection and to reduce pseudoreplication we screened the entire accessible area of each pond and collected no more than one egg per leaf or two eggs per plant. In marbled newts adult population size can be very small, especially in small water bodies such as springs (SCHOORL and ZUIDERWIJK, 1980; JEHLE et al., 2001; 2005) and this explains why sample size is small on some occasions (N<5 in four populations). Adults and larvae were captured by dip-netting. Tail tips were collected from adults and larvae were sacrificed. All samples were immediately stored in liquid nitrogen and later transferred to -80º C until the day they were analysed.

Genetic Data

All the tissue samples (from n≥10 individuals for most locales) were analyzed for four allozyme loci: peptidase A (Pep-A), peptidase B (Pep-B), peptidase D (Pep- D) and lactate dehydrogenase (Ldh-2; this locus is not yet expressed in embryos), that yield a species specific enzyme profile using standard starch gel

electrophoresis and isoelectric focusing. The genetic signature is consistent with morphological identification of adult marbled newts (CHAPTER 7).

With the program ARLEQUIN (Version 3.1; EXCOFFIER et al., 2005), we tested for departures from Hardy-Weinberg expectations and linkage

disequilibrium. We used the program FSTAT (v. 2.9.3.2; GOUDET, 1995) to calculate F-statistics and, to detect population sub-structuring, we analysed the results with STRUCTURE 2.1 (PRITCHARD et al., 2000). Using a Markov chain Monte Carlo (MCMC) algorithm, STRUCTURE assigns individuals to a population, or jointly to two or more populations, if their genotypes indicate that they are admixed. This is done assuming a model with K populations (where K may be

unknown), where each is characterized by a set of alleles for each locus.

Individuals are assigned to populations as to maximize Hardy-Weinberg and linkage equilibrium. We choose for the 'admixture model' because neighbouring populations may interbreed and varied K from 1 to 5 with 10 000 generations as the length of burn-in period and 100 000 MCMC generations after burn-in.

Convergence occurred before 10 000 in test runs. The value of λ was inferred by the program. We accepted K as indicating the true number of genetic partitions when the difference in likelihood between two consecutive values of K was maximal.

We investigated the existence of hybrids in our sample by performing a Bayesian model-based clustering as implemented in the software NEWHYBRIDS

(ANDERSON and THOMPSON, 2002). This computes the posterior probability that each individual belongs to each of six predetermined classes (pure species A, pure species B, F1-hybrid, F2-hybrid, backcross to species A and backcross to species B).

Ecological Data

For the spatial environmental analysis in a Geographical Information System (GIS), we selected 21 ecological parameters following Teixeira et al. (2001) and an additional explanatory variable (land surface occupied by orchards, arcsin transformed percentages – ORCH) that appeared locally informative. We used ORCH as a proxy to one or more unidentified variables that - possibly more directly than ORCH - would help to explain the newt distribution. The advantage of the parameter ORCH is that it manifests itself from field observations and that blanket data are readily available (INE, 1999).

For all variables, information was available in digital format for Portugal (DGA, 1995). A vegetation map (normalised difference vegetation index or NDVI) was obtained courtesy of the Royal Dutch Meteorological Institute (KNMI). An altitude map was taken from the internet

(http://edcwww.cr.usgs.gov/doc/edchome/ datasets/edcdata.html) and used to produce a relief map by a set of filter operations (ITC, 1997). Maps on the mean January and July temperature were digitalised from the Portuguese climate atlas (SCN, 1974).

A hierarchical clustering based on Spearman’s correlation coefficient was used to evaluate the level at which ecological information appeared redundant.

One variable was selected arbitrarily out of a set of variables that correlated at Spearman’s rs>0.8. Sixteen variables were retained: acidity of the soil (ACID), altitude , chlorates content of subterranean water (CHLO), frost days (FROD), frost months , hardness of subterranean water (HARD), humidity of the air (HUMI), insolation (INSO), lithology (LITH), vegetation index (NDVI), orchard land coverage , mean annual precipitation (PRET), relief (RELI), the sulphate content of subterranean water (SULP), mean annual temperature (TEMP) and mean July temperature (TJUL). To increase the comparability of their effects, all continuous variables were standardized to an average of zero and a standard deviation of one. The variables were introduced into the GIS analytical software as raster layers with 1 km spatial resolution. Values for 10*10 km UTM grids were obtained by averaging the data (modal values for the categorical variable LITH).

Model selection

We used an information-theoretic model selection approach for the statistical analysis of the data (BURNHAM and ANDERSON, 2002). Using our field knowledge, we built several a priori models that would explain the current distribution of the two species. By contrasting the presence of one species against the other, we circumvented the inclusion of false absence data. To understand the effect of the variable ORCH, we used our available data on fruit growing in the region as dependent variable, and applied a Stepwise Multiple Regression (SMR) with the same environmental data (all but ORCH) as explanatory variables, using the software SPSS v14 (SPSS, Inc., 2005). We then substituted ORCH by the model derived from this analysis in our a priori models, and added them to the list as a posteriori models. In a second step, we used the small-sample Akaike’s Information Criterion (AICc; BURNHAM and ANDERSON, 2002) to rank the models and chose the best one. AICc is defined as:

where ln L is the natural logarithm of the likelihood function, K is the number of parameters from the model, and n is the sample size. Akaike’s Information Criterion prevents overfitting the model, by taking the number of parameters into

consideration. To fit the models to the data and obtain the log-likelihoods of the models, we used the logistic regression procedure available in SPSS.

Results

We analyzed 398 individuals (101 adults, 50 larvae and 247 embryos; see allele frequencies in supplementary material – Table S1).

F-statistics show a high degree of population substructuring (Fst) in Pep-A (0.84) and Pep-D (0.66), intermediate in Ldh-2 (0.41) and low in Pep-B (0.14).

The exact test on Hardy-Weinberg expectations showed that population 4 presents a significant departure at one locus (Pep-D) under Bonferroni correction. Linkage disequilibrium test are significant at P<0.05 in two T. marmoratus populations that are neighbouring T. pygmaeus populations (population 23 and 31; see Figure 1 and Table S2 in supplementary information).

The model-based clustering method implemented in STRUCTURE 2.1 showed that the most likely number of partitions for the present data is K=2 (Table 1; for details see table S2 in supplementary material). The partitions correspond to T. marmoratus and T. pygmaeus. Since the likelihood continues to increase with K, we looked into the population structure when K=3, to examine the possibility that additional sub-structuring existed. The only result was that one of the clusters split into two equal parts. Fst values are consistent with the relative diagnostic power of the loci (high for Pep-A and Pep-D, intermediate for Ldh-2 and low for Pep-B). Accurate species identification is independent of the number of loci studied (Figure 1).

Figure 1 Proportion of membership of individuals to species 1 (= Triturus pygmaeus) using the software Structure 2.1 (Pritchard et al., 2000).

Data are separated out for samples with 0, 25, 50 and 75% missing data to show that, in practical terms, discriminatory power is independent of the loci used.

K logL SD

1 -1737,7 5,3

2 -961,9 0,6

3 -887,5 1,4

4 -849,4 30,7

5 -802,2 6,9

The spatial distribution of the two groups indicates the existence of a set of T. marmoratus populations surrounded by T. pygmaeus populations, forming an enclave in the centre of our study area (Figure 2; note that the isolated pocket is - in proper terms - an exclave of T. marmoratus and an enclave of T. pygmaeus).

Only population 18 had individuals of both species.

The Bayesian-based assignment of individuals to hybrid classes using NEWHYBRIDS indicates that hybridization, backcrossing and introgression are locally rare Thirty out of 31 populations were classified as either T. marmoratus or T. pygmaeus and not both. Population 18 contained both species as well as one individual with about equal probability of being a pure T. marmoratus (p=0.53) as having mixed species parentage (p=0.47, being either a F2-hybrid (P=0.35) or a backcross hybrid in direction of T. marmoratus (p=0.12)). This mixed population (population in grey in figure 2) was excluded from the spatial-environmental analysis.

Figure 2 Distribution of marbled newts in Portugal with Triturus marmoratus (hatched) and Triturus pygmaeus (shaded). The box shows the research area around Caldas da Rainha (a). The detailed map (b) shows the localities with T. marmoratus (solid dots) and T. pygmaeus (open dots).

See Table 1 for population details. Voronoi polygons are used to estimate the contiguous species distribution over the area. Note that the scale of extrapolation at the exterior of the study area is set

at c. 6 km as to match the level of interpolation.

Table 1 – Log-likelihood (logL) and standard deviation (SD) of the number of partitions (K) in our data set. K varied from 1 to 5. The increase in likelihood is not significant other than at K=2.

Model-2 ln LKnAICcHypothesis ORCHRELITJULFRODALTIPRETHARDHUMINDVITEMPSULPACIDConstant a priori 125.5013053.14Extent of orchard plantations2.85-1.23 236.8713075.88Relief-1.04-0.12 340.2413082.62July temperature-0.420.05 441.2313084.60Number of frost days-0.040.08 524.8523054.15Combination of model 1 and model 22.57-0.61-1.24 624.0233054.96Model 5 plus altitude2.78-0.33-0.83-1.49 724.7433056.40Model 5 plus precipitation2.50-0.56-0.22-1.21 824.7733056.47Combination of model 4 and model 52.62-0.57-0.18-1.25 920.7233048.36Combination of model 3 and model 55.10-0.40-2.10-3.30 1028.5263072.70from Arntzen (unpublished)-0.98-0.440.16-1.81-0.87-1.88-0.68 a posteriori 1128.7353069.97SMR model-0.92-3.07-2.880.79-0.38-0.72 1227.0663069.76Model 5 with ORCH replaced by SMR model-1.06-0.70-3.18-2.220.90-0.75-0.77 1326.1273071.32Model 7 with ORCH replaced by SMR model-0.740.00-1.63-3.69-2.700.24-0.48-0.90 1426.8073072.70Model 8 with ORCH replaced by SMR model-1.140.42-0.92-3.19-2.141.02-0.84-0.83 1522.8273064.75Model 9 with ORCH replaced by SMR model-2.17-2.82-4.53-2.072.24-0.31-2.01

Model equation

Table 2 List of a priori and a posteriori models used to determine the relationship between the distribution of Triturus marmoratus and Triturus pygmaeus in the Caldas da Rainha area, Portugal. The parameter values were obtained by logistic regression analysis of presence data in the area (by contrasting the presence of one species with the presence of the other). The log likelihood of the models (–2 ln L), number of parameters in each model (K), and the number of data points (n) were used to calculate the small-sample Akaike Information Criterion (AICc). The hypothesis behind each model is also listed. Lowest AICc value is highlighted in bold. SMR, stepwise multiple linear regression; ORCH, orchard land coverage; RELI, relief; TJUL, mean July temperature; FROD, frost days; ALTI, altitude; PRET, mean annual precipitation; HARD, hardness of subterranean water; HUMI, humidity of the air; NDVI, vegetation index; TEMP, mean annual temperature; SULP, sulphate content of subterranean water; ACID, acidity of the soil (details in Teixeira et al. 2001 and Arntzen, 2006).

GIS-model 9 (ORCH + TJUL + RELI; see table 3) showed the lowest AICc value. This model explains 50% of the total variance observed. The

presence of T. marmoratus relates to the higher abundance of orchards, lower July temperature and smoother relief than for T. pygmaeus. In the subsequent analysis, five environmental variables: ALTI, ACID, HUMI, SULP and TEMP explained the variable ORCH. This model explained 69% of the variance of ORCH. When, however, ORCH was replaced by this set of variables, the models did not perform as well (Table 2).

Discussion

One of us (JWA) first described the existence of T. marmoratus in the research area through a sporadic observation in March 1998. Our results confirm the presence of T. marmoratus in six populations outside its documented range in a pocket near Caldas da Rainha. This pocket of T. marmoratus is fully surrounded by populations of T. pygmaeus. Considering that i) the distance of ca. 10 km that separates it from the main T. marmoratus distribution exceeds the dispersal capability of large bodied newts (ARNTZEN and WALLIS, 1991; THIESMEIER and KUPFER, 2000) and ii) that the species are locally strongly parapatric, we conclude that the pocket equals to an enclave.

The variable ORCH has a stronger effect than its fellow explanatory variables in models 5-9 (Table 3). It would be inappropriate though to extrapolate any model with ORCH over wider areas because in Portugal extensive fruit growing is particular to the Caldas da Rainha region.

In the habitat preference model, we assume that strongly preferred habitat is of high quality for the species (RAILSBACK et al., 2003). So, if a model has a good fit, it is usually assumed that the species/habitat system is in equilibrium and that the species distribution will only change if the environment around also changes. This ignores, however, effects of life history and dispersal. Areas may be suitable but out-of-reach, due for example to unsuitable habitat in between

realized and prospective ranges.

The current distribution of marbled newts in central Portugal is best explained by T. pygmaeus moving north from its previous range and superseding T. marmoratus that only persisted in areas with ecological conditions more suitable for the species. Because T. marmoratus and T. pygmaeus occasionally

hybridize, this scenario can in principle be tested by looking for T. marmoratus

‘genetic footprints’ in areas that are hypothesized to have been inhabited by T.

marmoratus in the past (ARNTZEN and WALLIS, 1991). Similarly, we predict the absence of T. pygmaeus ‘genetic footprints’ inside the enclave.

There are several accounts of moving hybrid zones in, for example,

butterflies (MALLET et al., 1990; BLUM, 2002; DASMAHAPATRA et al., 2002), birds (PEARSON, 2000; ROHWER et al., 2001), crayfish (PERRY et al., 2001), lizards (HILLIS and SIMMONS, 1986) and amphibians (ARNTZEN, 1978; ARNTZEN and WALLIS, 1991); see also Barton & Hewitt (1985; 116-119). Most studies provide direct evidence of hybrid zone movement through the tracing of genetically interacting species distributions over time. Our study utilizes a single temporal window and provides compelling evidence for spatial change in a mosaic hybrid zone nevertheless, through the demonstration of an enclave.

The distance between T. marmoratus in the enclave and the main

distribution is minimally 6 km (the distance between populations 13 and 15) and maximally 15 km (the distance between populations 22 and 23). Although we cannot pertinently exclude the presence of long distance dispersal, such a scenario is unlikely given the absence of T. marmoratus or genetically mixed individuals in populations 13 and 15. Similarly, we cannot exclude the possibility that a human introduction is responsible for the enclave. There is, however, no tradition of newt husbandry in Portugal, and a deliberate or accidental release is improbable.

To infer the direction of the movement in a hybrid zone, it is equally possible to follow a direct or an indirect strategy. Direct demonstrations employ two or more temporally separated observations, either on position (e.g. HILLIS and SIMMONS, 1986 in Pholidobolus lizards) or shape of the cline that separates in this case connects the hybridizing species (e.g. DASMAHAPATRA et al., 2002 in Anartia butterflies). An indirect way is to look at disequilibrium measures. Cruzan (2005) showed in the wide Piriqueta coaroliniana/viridis (flowering plants from the family Turneracea) hybrid zone that P. viridis alleles were moving north. This was done by showing that the southern border of the hybrid zone presented relatively high levels of disequilibria, indicating recent gene flow from parental populations south of the hybrid zone. Interestingly, we observed significant levels of linkage disequilibrium in two T. marmoratus populations (populations 23 and 31) that are both within the dispersal range of T. pygmaeus populations. This strengthens the

argument that T. pygmaeus advances into T. marmoratus territory, even though we were unable to demonstrate current T. pygmaeus presence.

As a further test to our explanation, we predict that the presence of alien T.

marmoratus genes in T. pygmaeus exceeds that of the reverse condition, in a spatial pattern consistent with enclave formation.

Despite the range expansion of T. pygmaeus in its northern border, the situation in the south of Spain and Portugal is different. Due to desertification and an intensification of agricultural practices, T. pygmaeus is losing many breeding sites and has been classified as 'near threatened' (ARNTZEN et al., 2006). This is consistent with recent evidence that suggests that climate warming will not only increase the northern range of species but also decrease the southern one (THOMAS

et al., 2006).

If T. pygmaeus continues its competitive advance, the T. marmoratus enclave would eventually disappear. As yet, we have no indication on the speed of the process. Monitoring the area would provide an additional test to the hypothesis of T. pygmaeus expansion and document the speed of the process. It is remarkable that T. marmoratus is not only losing out to its sister-species at the southern edge of its range, but also to the related species T. cristatus at its northern edge. The advance of T. cristatus at the expense of T. marmoratus was estimated to occur at a speed of ca. one km per year (ARNTZEN and WALLIS, 1991). If range

replacement would proceed at this speed at either side of its range T. marmoratus would be squeezed out in ca. 500 years.

Acknowledgments.

We are grateful to Armando da Costa Pais from Direcção Regional da Agricultura do Alto Oeste for discussion on the relevance of agricultural data for our work.

The study was carried under license from the ICN (Instituto de Conservação da Natureza) in accordance to National law for capturing wild fauna and was financed by FCT (Fundação para a Ciência e Tecnologia) research project POCTI/34110/99.

References

ANDERSON,E.C. and E.A.THOMPSON (2002) A Model-Based Method for

Identifying Species Hybrids Using Multilocus Genetic Data. Genetics 160, 1217-1229.

ARNTZEN,J.W. (1978) Some Hypotheses on Postglacial Migrations of the Fire- Bellied Toad, Bombina bombina (Linnaeus) and the Yellow-Bellied Toad, Bombina variegata (Linnaeus). Journal of Biogeography 5, 339-345.

ARNTZEN,J.W. (2006) From descriptive to predictive distribution models: a working example with Iberian amphibians and reptiles. Frontiers in Zoology 3, 8.

ARNTZEN,J.W., P.BEJA, J.BOSCH, M.TEJEDO, M.LIZANA, R.JEHLE, I.

MARTÍNEZ-SOLANO, A.SALVADOR, M.GARCÍA-PARÍS, E.R.GIL, R.

MARQUEZ, P.SÁ-SOUSA and C.D.PANIAGUA (2006) Triturus pygmaeus.

In: IUCN 2008. 2008 IUCN Red List of Threatened Species.

www.iucnredlist.org. Acedido em 12 de Novembro de 2008.

ARNTZEN,J.W. and G.P.WALLIS (1991) Restricted gene flow in a moving hybrid zone of the newts Triturus cristatus and T. marmoratus in western France.

Evolution 45, 805-826.

ARNTZEN,J.W. and G.P.WALLIS (1999) Geographic variation and taxonomy of crested newts (Triturus cristatus supersecies): morphological and

mitochondrial data. Contributions to Zoology 68, 181-203.

BARTON,N.H. and G.M.HEWITT (1985) Analysis of Hybrid Zones. Annual Review of Ecology and Systematics 16, 113-148.

BIERNE,N., P.BORSA, C.DAGUIN, D.JOLLIVET, F.VIARD, F.BONHOMME and P.

DAVID (2003) Introgression patterns in the mosaic hybrid zone between Mytilus edulis and M. galloprovincialis. Molecular Ecology 12, 447-461.

BLUM,M.J. (2002) Rapid movement of a Heliconius hybrid zone: Evidence for Phase III of Wright's shifting balance theory? Evolution 56, 1992-1998.

BRIDLE,J.R., S.J.E.BAIRD and R.K.BUTLIN (2001) Spatial structure and habitat variation in a grasshopper hybrid zone. Evolution 55, 1832-1843.

BURNHAM,K.P. and D.R.ANDERSON (2002) Model Selection and Multimodel Inference: A Practical-Theoretic Approach, 2nd Edition edn. Springer, New York.

CRUZAN,M.B. (2005) Patterns of introgression across an expanding hybrid zone:

analysing historical patterns of gene flow using nonequilibrium approaches. New Phytologist 167, 267-278.

DASMAHAPATRA,K.K., M.J.BLUM, A.AIELLO, S.HACKWELL, N.DAVIES, E.P.

BERMINGHAM and J.MALLET (2002) Inferences from a rapidly moving hybrid zone. Evolution 56, 741-753.

DGA (1995) Atlas do Ambiente. Versão 1.1. Direcção Geral do Ambiente.

Ministério do Ambiente e dos Recursos Naturais., Lisboa, Portugal.

EXCOFFIER,L., G.LAVAL and S.SCHNEIDER (2005) Arlequin version 3.0: an integrated software package for population genetics data analysis.

Evolutionary Bioinformatics Online 1, 47-50.

GOUDET,J. (1995) FSTAT (Version 1.2): A computer program to calculate F- statistics. Journal of Heredity 86, 485.

HARRISON,R.G. (1990) Hybrid zones: Windows on evolutionary processes.

Oxford University Press, New York.

HEWITT,G.M. (1988) Hybrid zones-natural laboratories for evolutionary studies.

Trends in Ecology & Evolution 3, 158-167.

HILLIS,D.M. and J.E.SIMMONS (1986) Dynamic change of a zone of parapatry between two species of Pholidobolus (Sauria: Gymnophtalmidae). Journal of Herpetology 20, 85-87.

INE (1999) Recenseamento Geral da Agricultura (Available from http://www.ine.Part/prodserv/Rga/rga99_2.asp).

JAEGER,R.G. (1970) Potential Extinction Through Competition Between Two Species of Terrestrial Salamanders. Evolution 24, 632-642.

JAEGER,R.G. (1971) Competitive Exclusion as a Factor Influencing the

Distributions of Two Species of Terrestrial Salamanders. Ecology 52, 632- 637.

JEHLE,R., J.W.ARNTZEN, T.BURKE, P.KRUPA and W.HÖDL (2001) The annual number of breeding adults and the effective population size of syntopic newts (Triturus cristatus, T. marmoratus). Molecular Ecology 10, 839- 850.

JEHLE,R., G.A.WILSON, J.W.ARNTZEN and T.BURKE (2005) Contemporary gene flow and the spatio-temporal genetic structure of subdivided newt

populations (Triturus cristatus, T. marmoratus). Journal of Evolutionary Biology 18, 619-628.

JIGGINS,C.D. and J.MALLET (2000) Bimodal hybrid zones and speciation. Trends in Ecology and Evolution 15, 250-255.

KEY,K.H.L. (1981) Species, Parapatry, and the Morabine Grasshoppers.

Systematic Zoology 30, 425-458.

MALLET,J., N.BARTON, L.M.GERARDO, S.C.JOSE, M.M.MANUEL and H.

EELEY (1990) Estimates of Selection and Gene Flow From Measures of Cline Width and Linkage Disequilibrium in Heliconius Hybrid Zones.

Genetics 124, 921-936.

OBSERVATION,I.I. F.G.-I.S. A.E. (1997) Ilwis 2.1 for Windows. ITC, Enschede, the Netherlands.

PEARSON,S.F. (2000) Behavioral asymmetries in a moving hybrid zone. Behav.

Ecol. 11, 84-92.

PERRY,W.L., J.L.FEDER, G.DWYER and D.M.LODGE (2001) Hybrid zone dynamics and species replacement between orconectes crayfishes in a northern Wisconsin lake. Evolution 55, 1153-1166.

PRITCHARD,J.K., M.STEPHENS and P.DONNELLY (2000) Inference of Population Structure Using Multilocus Genotype Data. Genetics 155, 945-959.

RAILSBACK,S.F., H.B.STAUFFER and B.C.HARVEY (2003) What can habitat preference models tell us? Tests using a virtual trout population.

Ecological Applications 13, 1580-1594.

RAND,D.M. and R.G.HARRISON (1989) Ecological genetics of a mosaic hybrid zone: Mitochondrial, nuclear, and reproductive differentiation of crickets by soil type. Evolution 43, 432-449.

ROHWER,S., E.BERMINGHAM and C.WOOD (2001) Plumage and mitochondrial DNA haplotype variation across a moving hybrid zone. Evolution 55, 405- 422.

SCHOORL,J. and A.ZUIDERWIJK (1980) Ecological Isolation in Triturus cristatus and Triturus marmoratus (Amphibia: Salamandridae). Amphibia-Reptilia 1, 235-252.

SCN (1974) Atlas climatológico de Portugal Serviço Climático Nacional, Lisboa, Portugal.

SITES,J.W.J., M.MORANDO, R.HIGHTON, F.HUBER and R.E.JUNG (2004) Phylogenetic Relationships of the Endangered Shenandoah Salamander (Plethodon shenandoah) and Other Salamanders of the Plethodon cinereus Group (Caudata: Plethodontidae). Journal of Herpetology 38, 96-105.

SPSS,I. (2005) SPSS for Windows. SPSS Inc., Chicago.

TEIXEIRA,J., N.FERRAND and J.W.ARNTZEN (2001) Biogeography of the golden- striped salamander Chioglossa lusitanica: a field survey and spatial modelling approach. Ecography 24, 618-624.

THIESMEIER,B. and A.KUPFER (2000) Der kammolch — ein wasserdrache in gefahr. Beiheft 1 der Zeitschrift für Feldherpetologie. Laurenti-Verlag, Berlin.

THOMAS,C.D., A.M.A.FRANCO and J.K.HILL (2006) Range retractions and extinction in the face of climate warming. Trends in Ecology & Evolution 21, 415-416.

Appendix S1

ation12345678910111213141516172627282930181920212223242531 Adults02020220601118101000000100701102000 Larvae50006000000010030000009012302000 Eggs10050081110013008210011510101020211401413182021 Total1520252261411111148101021031151011102162116241417182021 es Pep-A N13191222610101011071092103100101010291010221010111010 a0,560,850,850,910,950,951,001,000,60 b0,050,05 d1,001,001,001,001,001,001,001,001,001,001,001,001,001,001,001,001,001,001,001,001,000,440,150,150,050,40 e0,05 Pep-B N1520252261310101118101028345020116201424141411119 b0,130,240,070,080,230,060,150,170,090,030,060,06 d0,370,780,540,660,920,690,801,001,000,950,880,600,951,000,810,831,000,701,001,000,840,950,960,901,000,960,910,770,89 e0,630,100,220,270,080,200,000,000,050,060,250,050,190,300,060,030,040,040,040,090,230,06 Pep-D N1520242261291011271082831150121162015231415112011 f0,970,850,690,700,920,710,940,501,000,331,001,000,941,000,810,831,000,901,001,001,000,130,03 d0,020,080,060,170,631,001,000,831,000,871,001,001,00 e0,100,020,180,170,67 g0,030,050,110,060,060,130,10 c0,020,500,030,03 a0,090,170,07 b0,13 h0,250,13 Ldh-2 N5202022660001810200300010014012303000 c0,03 b1,001,000,981,001,001,001,001,001,001,000,831,000,460,500,430,67 f0,170,540,500,57 e0,33

Triturus pygmaeusTriturus marmoratus

ndix S1. Allele frequencies in XLS format. Allele frequencies over four loci in marbled newts from the Caldas da Rainha area, Portugal. lysis of the results indicates the existence of two separate genetic units that correspond to T. pygmaeus (22 populations) and T. marmoratus lleles indication are as in Espregueira Themudo & Arntzen (2007). N is sample number; zero sample size fers to missing data.

Appendix S2

Clustering assignment of 31 sample sites in XLS format. Assignment of 31 populations of marbled newts (Tritrus marmoratus and T. pygmaeus) to two clusters in STRUCTURE 2.1 . Columns represent sample sites localities, Universal Transverse Mercator coordinates (UTM_X and UTM_Y), sample number (N) and the probability to belong to cluster 1 (that corresponds to T.

marmoratus).

Code Sample site UTM_X (km) UTM_Y (km) N Inferred cluster 1

1 Porto de Mós 514.0 4384.3 15 0.009

2 Rio Maior 507.3 4355.2 20 0.008

3 Alqueidão 536.0 4376.6 25 0.020

4 Valado dos Frades 499.2 4383.1 22 0.007

5 São Bartolomeu dos Galegos 476.0 4348.0 6 0.044

6 Mosteiro de Alcanene 513.7 4364.5 14 0.014

7 Foz do Arelho 482.6 4365.0 11 0.016

8 Casais dos Morgados 488.6 4369.5 11 0.015

9 Carrascal 503.7 4375.9 1 0.010

10 Molianos 506.8 4373.0 14 0.016

11 Covas 517.5 4376.1 8 0.020

12 Casais Monizes 509.0 4367.0 10 0.007 Probability

13 Carrascal II 503.8 4375.8 10 0.020 Sample Inferred cluster 1

14 Cela 496.3 4376.6 2 0.011

15 Genrinhas 498.4 4373.9 10 0.037 J1 0.035

16 Sta Susana 499.3 4353.0 3 0.119 J2 0.036

17 Pataias Gare 501.5 4388.6 11 0.011 J3 0.226

18 Famalicão da Nazaré 492.5 4377.3 5 0.018 J4 0.631

19 Chão 516.4 4378.8 10 0.032 J5 0.870

20 Covão da Fonte 517.6 4376.3 11 0.031 J6 0.938

21 Molianos II 508.2 4374.4 10 0.029 J7 0.941

22 Ribeira da Maceira 500.4 4371.4 2 0.027 J8 0.942

23 Juncal 507.3 4383.4 16 0.778 J9 0.955

24 Casal da Charneca 491.5 4360.0 21 0.984 J10 0.974

25 Andam 508.9 4385.8 16 0.974 J11 0.974

26 Salir de Matos 493.0 4364.4 24 0.987 J12 0.977

27 Cós 504.4 4383.8 14 0.990 J13 0.987

28 Fonte da Pena da Couvinha 497.5 4370.6 17 0.971 J14 0.987

29 Vidais 495.2 4358.3 18 0.984 J15 0.987

30 Casal da Coita 500.5 4366.5 20 0.986 J16 0.992

31 Vimeiro 498.6 4368.8 21 0.934